Sorption behavior of malachite green onto pristine lignin to evaluate the possibility as a dye adsorbent by lignin

The objective of this study was to evaluate the adsorption characteristics of malachite green (MG) on pristine lignin as a dye adsorbent. The adsorption capacity of MG on lignin (31.2 mg/g) was described by Langmuir isotherm and pseudo second order models, and were higher than humic acid (6.4 mg/g). The adsorption of MG by lignin was rapid occurring within 15 min of the reaction, and then equilibrium was reached. The adsorption of MG by lignin based on an intraparticle diffusion model indicated that it was dominated by external boundary. Removal of MG by lignin can be applied at a wide range of pH’s (2–5), and optimal lignin dosage for MG removal was 3 g/L. In addition, the desorption efficiency of MG adsorbed on lignin was highest in methanol + acetic acid (95:5%, v/v) mixture of all solutions tested. The peaks attributed to the hydrogen-bonded stretching vibrations and sulphonyl groups in lignin before MG adsorption, were assigned at about 3400 and 620 cm−1, while the peaks in lignin after MG adsorption were attenuated or reduced. This result indicates that the adsorption of MG by lignin is closely related to the O–H and S–O bonds. Finally, this study suggests that pure lignin, which excludes active processes, can also be used as an adsorbent for dyes. However, in order to utilize the dye-adsorbed lignin repeatedly, further studies will be needed.


Introduction
Recently, research into converting woody biomass into bioethanol and chemical materials has attracted much attention, and has already reached the commercialization stage [1]. In particular, the amount of lignin as a by-product of manufacturing processes increases with increasing bioethanol production, which has now reached 50 million tons per year [2]. Considering the potential for growth of the bioethanol market in future, lignin discharges are expected to increase further. Lignin is an amorphous material with an aromatic structure, which accounts for about 25-35% of woody biomass, the second most abundant natural polymeric substance after cellulose [3]. Lignin is composed of a monomer of a phenylpropane (C6-C3) structure in which a phenolic hydroxyl group and a methoxy group are bonded, and typically has a molecular structure in which three kinds of phenylpropane monomers (para-coumaryl alcohol, coniferyl alcohol and synapyl alcohol) are crosslinked [4,5].
However, lignin is less useful due to this complicated structure, and only about 2% is used as a dispersant, an adhesive, or a surfactant, and most of the lignin produced as a by-product is discarded or incinerated [6]. Most recently, Yin et al. [7] reported that lignin has a number of functional groups including hydroxyl, methoxyl and carbonyl groups, and thus can be utilized as a raw material for chemical manufacture through chemical modification techniques such as oxypropylation and epoxidation [8]. Some researchers have found that materials derived from the chemical modification of lignin can be utilized as raw materials for making plastics such as polyurethanes and polyesters in the manufacture of plastics [9,10], as well as raw materials for phenolic resins, epoxy resins and carbon fiber products. In terms of agriculture, lignin has a similar structure to humic acid, so that it is possible to utilize humified lignin as an agricultural fertilizer through chemical and biological modification processes [11]. However, lignin modification involves some technically demanding approaches and significant labor input, as well as coming with high costs. Furthermore, considering the daily production levels of lignin, the ability to recycle lignin by these modification methods is limited. In order to overcome these problems, there is a pressing need for a recycling method capable of large-scale application which utilizes the chemical and physical characteristics of lignin itself, without the need for modification.
As mentioned above, lignin has an aromatic threedimensional polymer structure containing functional groups such as phenol, hydroxyl, carboxyl, methoxyl and aldehyde groups. These functional groups have been reported by many researchers to be quite effective in removing cationic contaminants including heavy metals and dyes in wastewater [12]. For example, some studies found that lignin was effective for removing heavy metals [13][14][15]. On the other hand, the adsorption of dyes by lignin was mostly evaluated by using activated lignin subjected to acid/alkali treatment, metal impregnation and pyrolysis [16][17][18][19]. Information about the adsorption behavior of dyes on pristine lignin however, is extremely limited. In particular, no studies have investigated the use of lignin as an adsorbent for the removal of malachite green (MG) from water.
Malachite green is a distinctive example of a basic and cationic dye, that has been widely used in medicine, fisheries, food and directly as a dye in the wood, silk, leather and paper industries. However, high concentrations of MG are toxic if discharged into the aqueous environment, with carcinogenic effects on human beings and causing suffocation of aquatic plants [20].
Therefore, the objective of this study was to evaluate the adsorption characteristics of MG by pristine lignin in order to determine whether it may be an appropriate strategy for increasing the recyclability of lignin in a way that utilize its own characteristics.

Materials
Lignin, humic acid and MG were purchased from Sigma-Aldrich (Gillingham, UK). An MG stock solution was prepared at 1000 mg/L using distilled water and diluted to the concentration desired for each experiment. Solutions were adjusted to pH 5 using 0.1 M HCl or NaOH, except in the case of determining the effect of pH on MG adsorption.

Methods
Determination of the maximum adsorption capacities for MG by lignin and humic acid were performed at different initial MG concentrations. In brief, 25 mL of MG diluted solution ranging from 5 to 300 mg/L were added to a set of 100 mL glass flasks containing 0.05 g of either lignin or humic acid. After incubation with shaking for 24 h at 25 °C, the mixed samples were separated by centrifugation at 5000 rpm for 10 min. The concentrations of MG in residual solutions were analyzed by UV-Vis spectrometer (UV-1800, Shimadzu, Japan). The amount of MG adsorbed per mass unit of adsorbent was calculated by difference between the initial and equilibrium concentrations in solution. In order to determine the maximum adsorption capacity, the amount of MG adsorbed by lignin and humic acid based on initial MG concentration was applied to Freundlich and Langmuir isotherms. The equations, plots and relationship constants for the Fruendlich and Langmuir isotherms are given in Additional file 1: Table S1.
The dynamic adsorptive behavior of the adsorbent is one of the factors that must be considered for process design and operation control of the adsorption system. In this study, the dynamic adsorptive behavior was evaluated by weighing 0.05 g of lignin and humic acid in Erlenmeyer flasks (100 mL) followed by 25 mL additions of MG (50 mg/L). Samples were then equilibrated at the designated time intervals of up to 24 h. The separation and analysis of the stirred samples were carried out in the same manner as described in the isotherm experiment. To understand the adsorption mechanisms of MG by lignin and humic acid, pseudo-first-order (PFO), pseudosecond-order (PSO) and intraparticle diffusion (IPD) models were used to investigate the adsorption kinetics (Additional file 1: Table S1). In addition, the adsorption properties of MG by lignin and humic acid were evaluated at different initial pH (2-7; 50 mg/L MG) and dosage (1-4 g/L; 100 mg/L MG).
The desorption efficiency of MG-adsorbed lignin and humic acid, were evaluated in the presence of different solvents; methanol (MeOH), methanol + acetic acid (95:5%, v/v; MeOH-Ac) mixture and diluted acetic acid (5% in distilled water, v/v; Ac). MG-adsorbed samples were washed with distilled water and dried. Desorption solutions (25 mL) were added to a set of glass flasks containing 0.05 g of the MG-adsorbed sample, followed by stirring for 4 h. All analytical methods are the same as those above. The recyclability of lignin was subsequently evaluated, using the optimal desorption solution (MeOH-Ac), in desorption experiments.
Finally, we characterized the involvement of functional groups on lignin before and after MG adsorption, by Fourier transform infrared spectrometer (FTIR, Bruker VER-TEX 70, Germany) in the range 400-4000 cm −1 , to define more precisely the chemical interactions between adsorbent and MG.

Isotherm
The adsorption characteristics of MG by lignin and humic acid at different initial dye concentrations is shown in Fig. 1 and Table 1. As the solution concentration of MG exposed to both lignin and humic acid increased, the adsorbed amount progressively increased, whereas it reached an equilibrium at MG concentration of 100 mg/L (Fig. 1a). The amount of MG adsorbed by lignin and humic acid under these conditions was applied to Langmuir and Freundlich isotherm models. The results demonstrated that the amount of adsorbed MG was better represented by the Langmuir model (R 2 = 0.9951-0.9995) than the Freundlich model (R 2 = 0.8514-0.9349) (Fig. 1b, c. The Freundlich model assumes adsorption by active sites in heterogeneous surfaces, whereas the Langmuir model assumes adsorption by a monolayer in homogeneous surfaces [21]. Therefore, the adsorption of MG by lignin and humic acid was mechanistically dominated by monolayer adsorption to a homogeneous surface. Of particular note, the maximum adsorption capacity (a) of MG by lignin was 31.2 mg/g, which was 4.9 times higher than that of humic acid (6.4 mg/g). The b value obtained from the Langmuir isotherm model shows the affinity between adsorbent and adsorbate, which was 2.08 for lignin and 0.17 for humic acid. These data indicate that lignin has a much higher adsorption affinity for MG than humic acid.
The maximum adsorption capacity of MG by lignin was compared to MG interactions with other adsorbents, which have previously been reported since no studies have been conducted on this specific topic.
As shown in Table 2, it is difficult to directly compare the amount of MG adsorbed by these materials, because experimental conditions, including initial MG concentration, dosage and pH used in other studies, were different. In recent times, most carbon-based byproducts have focused on the development of activated adsorbents through pyrolysis or chemical activation, and the adsorbed amounts of MG by them ranged from 12 to 435 mg/g [22][23][24]. The amounts of MG adsorbed by degreased coffee bean [25], oil palm trunk fiber [26] and rattan sawdust [27] were 55, 149 and 63 mg/g,  respectively, all of which were higher than that of lignin. Differences in amounts of MG adsorbed on these adsorbents is due to the different chemical and physical properties of the raw materials and the discharge route of byproducts.
In contrast to this underperformance by lignin, the amount of MG adsorbed by lignin was higher than that of bentonite and clay soil [28,29], but lower than that of zeolite [30]. Based on these comparisons it appears that lignin can be used as an adsorbent for MG removal from aqueous solution, however, the performance overall was inferior to that of other adsorbents previously reported. Therefore, it appears that a new activating technique that can remove various pollutants is needed, while continuing to emphasize lignin's inherent characteristics. Figure 2 shows the kinetic behavior of MG adsorption by lignin and humic acid with respect to reaction time. The amount of MG adsorbed increased rapidly up to 15 min and reached near equilibrium after 30 min (Fig. 2a). Generally, due to the abundant vacant sites in the fresh adsorbent, rapid adsorption occurs at the beginning of the reaction, and persists until the adsorption site becomes saturated [38,39]. Our study suggests that the active sites of lignin and humic acid for MG adsorption are mostly saturated within 15 min.

Kinetics
The dynamic adsorption behavior of MG on lignin and humic acid was evaluated by the general kinetic models PFO (Fig. 2b) and PSO (Fig. 2c). The constants (k 1 , k 2 and q e ) with correlation coefficient (R 2 ) obtained from these kinetic models are shown in Table 2. The adsorption of MG by lignin and humic acid was better described by the PSO model (R 2 = 0.3403 for humic acid and 0.6901 for lignin) than by the PFO model (R 2 = 0.9982 for humic acid and 1.000 for lignin), which indicate that MG adsorption by lignin and humic acid is controlled by chemosorption. This result suggested that MG adsorption by lignin and HA was a rate-limiting step and that the reaction rate is proportional to the active site numbers existing in lignin and humic acid [21,40]. In particular, the q e value derived from PSO of lignin was 3.2 times higher than that of humic acid, indicating that lignin had more active sites for MG adsorption than humic acid ( Table 3).
The IDP model is an important factor for determining whether adsorption of adsorbate by adsorbent is due to external boundary or internal (intraparticle) diffusion [41]. When the adsorption results of MG by lignin and humic acid were applied to the IPD model, the reaction rate was divided into two stages. The first straight line represents the adsorption by the outer boundary, and the second following line represents the adsorption by the internal diffusion model. Therefore, the results of the IPD on lignin and humic acid in this study, were derived only from the second straight line. In our study, the q i values determined from the IPD model were 0.34 for lignin and 0.41 for humic acid, which are lower than those for carbon adsorbents such as activated charcoal and biochar [37,42]. Therefore, adsorption of MG by lignin and humic acid is mostly dominated by external boundary.

pH effect
The adsorption characteristics of MG by lignin and humic acid based on initial pH are shown in Fig. 3. The MG standard solution (50 mg/L) was stable without any significant differences up to pH 5 (from 45.6 to 49.6 mg/L), but the absorption in UV-Vis spectrometer was reduced at higher pH values because the color of MG at the same wavelength, faded above pH 6. At pH 8, the  concentration of MG could not be measured because the color was completely faded. Tang et al. [18] also reported a similar result. The concentration of residual MG following adsorption by lignin was 6.8 mg/L at pH 2 but ranged between 1.3 mg/L and 2.4 mg/L for the rest of the pH value. On the other hand, when the amount of MG adsorbed by lignin was calculated based on the residual and the initial concentrations at each pH, the results showed 20.5, 23.3, 23.8, 24.2, 12.6, and 3.8 mg/g at pH 2, 3, 4, 5, 6, and 7, respectively. In order to determine the adsorbed amount of MG by lignin corresponding to the pH change, it is recommended that the pH should be adjusted to the range of 2-5.
The adsorption of MG by lignin increased with increasing pH from 2 to 5. In the acidic solution, the functional groups on the surface of the adsorbent bind H + and from positively charged protons. In addition, Harmita et al. [43] reported that the zeta potential charge of lignin is negative at pH 2. For these clear reasons, the adsorption of MG by lignin in the acidic state is expected to be at it's lowest rate. As the pH increases, hydrogen ion concentration in the solution reduces and MG is protonated increasing the amount of the dye adsorbed due to electrostatic attractions between lignin and MG.
In the adsorption experiment with humic acid, the adsorption capacity decreased as the pH increased from 2 to 5, a stark contrast to what was observed with lignin. Also, it appears that the adsorption capacity of MG increased at pH 6, however, this might have been due to the fading of MG standard solution. Gűrses et al. [44] reported that the increasing adsorption under acidic conditions may be related to preference of the cationic dyes for active sites and/or the increasing accessibility to interlayer regions of protonated (MGH 2+ ) and monomeric species, because of removal of some oxides at the adsorbent surface.
In particular it is worth noting that the residual MG concentration after adsorption of MG by humic acid at pH 7 was higher than the initial concentration. As shown in Fig. 3b, the final pH was lowered to less than 7 because of the pH buffering action of humic acid in solution, which is closely related to recovery of the faded color.

Dosage effect
The responses to lignin and humic acid doses are shown in Fig. 4. With increasing dosage the removal efficiency of MG increased (Fig. 4a). This is clearly due to the increasing number of active sites for adsorption. Additionally, the amount of MG adsorbed by lignin was highest at 3 g/L among all dosages tested (Fig. 4b). The increase in adsorption efficiency with dosage was more likely a function of increasing total amount of active sites [45], and our results indicate that not all of the newly available active sites are used for MG binding. For example, with humic acid, the adsorbed amount ranged between 6.24 and 6.51 mg/g regardless of the dose, indicating that not all the increased number of active sites contributed to the adsorption of MG.

Desorption
The DW, MeOH, MeOH-Ac and diluted-Ac solutions were used to desorb lignin-adsorbed MG (Fig. 5). On a technical note, the amount of humic acid remaining after the first adsorption reaction was reduced to about 33% of the initial dose, indicating that most of it was dissolved. Therefore, experiments for desorbing on humic acid could not be carried out. The desorption efficiency of DW on the MG adsorbed-lignin was negligible. On the other hand, the desorption efficiencies of MG-adsorbed lignin by MeOH, MeOH-Ac and diluted-Ac solutions were 28.9, 78.5 and 12.3%, respectively. Hence, the desorption efficiency of MG by MeOH-Ac was dramatically higher than any of the other solutions. This result indicates that MG was adsorbed by heterogeneous adsorption site mechanisms, which likely included electrostatic attraction, hydrophobic interaction and H-bonding [46]. Of interest, Afkhami et al. [47] reported that most of the dye adsorbed on surfaces, desorbed within 2 min if there was no internal diffusion resistance to the dye inside the adsorbent. We specifically looked at this possibility and tested desorption at 2 min and 24 h. The desorption efficiency in term of the percentage of total amount adsorbed was similar regardless of desorption time (data not shown). This may be further evidence that the adsorption of MG by lignin is not affected by internal diffusion, which is in agreement with the kinetic study and IPD model mentioned.
The recyclability of lignin was evaluated by initial adsorption followed by optimal desorption (MeOH-Ac). The amount of desorbed MG decreased as lignin was repeatedly re-used. However, more importantly, the amount of MG adsorbed by lignin in the second adsorption reaction was 2.3 mg/g, which was 84% lower than that of the first adsorption reaction. A reasonable explanation for these results is that the MG is not readsorbed following desorption because the C-OH functional groups on the lignin surface have been esterified by MeOH-Ac.
Baek et al. [25] reported that the desorption efficiency of MG adsorbed on degreased coffee bean using HCl decreased from 28.3 to 9.4% as the number of reactions increased (from 1 st reaction to 3 rd reaction). This study also investigated the desorption efficiency of MG adsorbed on lignin by varying HCl concentration. As the concentration of HCl increased, the desorption efficiency increased, but the amount of MG desorbed by 0.1 M HCl was approximately 2.3 mg/g, which was minimal as compared with the total adsorbed MG. The desorption efficiency of MG adsorbed on lignin under the same desorption solution (0.1 M HCl) was lower than that of the coffee bean, indicating that lignin has a higher affinity for MG than coffee bean.
Reversibility of adsorption is determined by whether there is strong bonding (ionic or/and covalent bond) or weak bonding (like Van der Waals' forces or/and dipoledipole interaction) between the adsorbent surface and the dye molecule [48]. Based on our results, it is apparent that the bonds between lignin and MG are strong and therefore difficult to desorb by any solution.

Functional group variation in lignin before and after MG adsorption
The FTIR spectra for lignin before and after MG adsorption are shown in Fig. 6. The peaks attributed to the hydrogen-bonded stretching vibrations in the lignin before MG adsorption were broadly assigned at about 3400 cm −1 [49], while the peaks in lignin after MG adsorption were attenuated or reduced.
In general, the peaks represented at 620 cm −1 in lignin were reported to be related to the sulphonic groups [50].